Knowledge and Attitudes of Croatian Nurses Toward Hypoglycemia Management: A Cross-Sectional Study

Abstract

Background/Objectives: Hypoglycemia remains the most frequent acute complication of diabetes, particularly among insulin-treated patients, with significant implications for morbidity, length of hospital stay, and healthcare costs. Nurses play a critical frontline role in its recognition and management, yet their competence varies widely. This study aimed to assess the knowledge and attitudes of Croatian nurses regarding hypoglycemia management and to identify key demographic and professional predictors. Methods: We conducted a cross-sectional online survey following CHERRIES guidelines of 317 nurses across Croatia using a validated 26-item knowledge test and a 6-item attitude scale. Descriptive statistics, Mann–Whitney U tests, and standardized effect sizes were used to assess group differences. Multivariable logistic and linear regression models examined the independent effects of education, sex, experience, and workplace setting. Results: The mean knowledge score was 66.9% (SD = 17.8), and the mean attitude score was 3.42 (SD = 0.70) on a 5-point scale. Nurses with tertiary education had significantly higher odds of achieving adequate knowledge (OR = 68.3, 95% CI: 19.9–234.2) and more favorable attitudes (β = +1.02, p < 0.001). Female sex had a small independent effect on knowledge (OR = 2.59, 95% CI: 1.02–6.62), while experience and workplace setting were not significant predictors. Conclusions: Although overall knowledge and attitudes were moderately positive, substantial disparities persist, particularly across educational levels. Clinical Practice Implications: These findings support integrating structured hypoglycemia training into nursing curricula and in-service programs to improve patient safety.

1. Introduction

Diabetes is a growing public health problem worldwide and in Croatia. Global diabetes prevalence is estimated at 9.8% [1], while in Croatia, it is approximately 6.9% [1]. Inpatient hypoglycemia affects up to 21.5% of hospitalized diabetic patients [2].

As the most frequent acute complication of diabetes, hypoglycemia threatens patient safety, contributes to prolonged hospital stays, increases healthcare costs, and raises mortality risk, particularly in insulin-treated patients [3].

While timely recognition and management of hypoglycemia is essential, several international studies have documented persistent gaps in nurses’ ability to respond to such events [4]. Nurses are often the first to detect clinical signs of deterioration, yet studies show substantial deficits in their ability to manage hypoglycemia in a timely and evidence-based manner [5,6,7,8].

Studies from diverse healthcare systems—including Iraq [9], Saudi Arabia [10], Jordan [11], the United Kingdom [12], and the Philippines [13]—have reported moderate-to-low levels of objective knowledge and inconsistent application of evidence-based practices. One study in particular found a striking disparity between perceived and actual diabetes knowledge among nurses (mean self-reported knowledge score = 46.9/60 vs. objective score = 25.4/49), highlighting a recurring overestimation of competence [10].

Several predictors of improved knowledge have been explored across studies, including higher educational attainment, perceived clinical competence, and access to institutional guidelines or diabetes-specific training [8,14,15]. Despite this, very few studies have employed multivariable statistical models to control confounding factors or to explore the independent contributions of variables such as experience, gender, or workplace type [16].

Community-based nurse-led interventions have shown significant effectiveness in improving outcomes for individuals with type 2 diabetes. Multiple studies have demonstrated reductions in HbA1c levels through various nursing interventions [17,18,19,20,21,22]. These interventions include self-management education programs, telephone coaching, and home visits, which have resulted in improved glycemic control, self-efficacy, and diabetes self-management behaviors [17,22,23]. The duration of interventions ranged from 3 to 15 months, with positive effects often sustained beyond the intervention period [17,19,21]. Community health workers and peer leaders have also been effective in maintaining improvements in HbA1c and other outcomes [21]. These community-based approaches have demonstrated particular benefits for underserved populations, including rural and low-income adults.

Targeted nurse-managed hypoglycemia protocols have markedly reduced inpatient events. Marelli et al. [24] demonstrated that a carbohydrate-focused treatment algorithm cut hypoglycemia incidence by over 30%, while Shea et al. [15] showed that a nurse-driven root cause analysis in critical care halved hypoglycemia rates. Compton et al. [25] further found that a nurse-directed blood glucose management protocol reduced low-glucose episodes without increasing workload.

Although Croatia faces a high burden of diabetes, the Croatian nursing system is stratified into three major educational levels: secondary vocational training, higher vocational programs, and university-level degrees.

International guidelines, such as those by the American Diabetes Association, emphasize the importance of standardized protocols for hypoglycemia management, underscoring the critical role of nursing preparedness [26].

The primary objective of this study was to assess Croatian nurses’ knowledge of hypoglycemia; secondary objectives were to evaluate attitudes and explore associations with experience and education. By doing so, this study contributes novel, actionable evidence to guide nursing education and improve inpatient diabetes safety.

2. Materials and Methods

2.1. Study Objectives

Specifically, we aimed to achieve the following:

Assess nurses’ knowledge and attitudes about hypoglycemia management using validated instruments.

Identify the independent effects of education, experience, sex, and workplace setting on these knowledge and attitudes outcomes.

2.2. Study Design

A cross-sectional study among nurses in Croatia was conducted in November and December 2023 using the “Questionnaire for nursing staff on inpatient hypoglycemia (low blood glucose) management”. The questionnaire was originally developed by Isnani et al.; it has been validated and used in previous studies [13].

The 26-item knowledge test and 6-item attitude scale were adapted from Isnani et al. [13] with established validity. A summary of all the items by domain is provided in Supplementary Table S1.

We followed the CHERRIES checklist for Internet surveys (Supplementary Table S2) and STROBE guidelines for cross-sectional studies

2.3. Ethics

Approval for this study was obtained from the Ethics Committee of the Croatian Catholic University (602-04/23-11/042 498-15-06-23-001).

2.4. Participants

The research targeted nurses in Croatia. Information about the study was provided before the questionnaire was provided, and the willingness to participate was considered as consent.

Invitations were posted in a Croatian Nurses Facebook group (≈18,700 members) from 10 November 2023 to 15 December 2023 via LimeSurvey.

Eligibility: Registered nurses across all departments, any career length, and educational level. LimeSurvey cookies and IP checks prevented duplicate entries; the logs were not retained post-analysis.

Scoring:

Knowledge: Each correct item scored 1; the total was converted to % (0–100).

Attitude: Six Likert items (1–5), two reverse-coded, averaged for a composite score.

2.5. Statistical Analysis

The survey responses were summarized using descriptive statistics, with categorical variables reported as absolute and relative frequencies, and continuous variables presented as medians with interquartile ranges (IQR), as well as minimum and maximum values.

Knowledge was assessed using 26 scored items based on multiple-choice and multiple-response questions, each contributing equally to a final score scaled from 0 to 100. Attitude was assessed using six 5-point Likert-scale items, with two reverse-coded statements, and an average score was calculated for each participant [13]. Internal consistency for the attitude scale was acceptable (Cronbach’s alpha = 0.65).

The homogeneity of the nurse subgroups based on educational level, age, total length of service, and tenure in healthcare was evaluated using the chi-square test. To test our hypotheses, the Mann–Whitney U test was employed, using measures of knowledge about hypoglycemia and attitudes toward its management as dependent variables and nursing professional education and healthcare experience as independent variables. This test was selected due to the non-normal distribution of the knowledge and attitude scores, confirmed through Shapiro–Wilk tests. Pearson correlation coefficients (r) were calculated to assess the linear relationships between percent-correct knowledge, mean attitude scores, and years of experience. All the statistical analyses were conducted using STATISTICA 12 (Tibco, CA, USA).

3. Results

3.1. Baseline Characteristics

A total of 317 respondents participated in the study, of whom 256 (80.8%) were female and 61 (19.2%) were male. Most participants reported working in tertiary (n = 135) or primary (n = 110) healthcare institutions, with smaller numbers in secondary care (n = 34), educational institutions (n = 30), or currently unemployed (n = 8).

Regarding education, 100 respondents (31.5%) held a secondary vocational diploma, 145 (45.7%) completed a higher vocational program, and 72 (22.7%) held a university-level nursing degree. For analysis, we grouped the latter two as “tertiary education” in line with Croatian and EU classification systems.

Most participants (n = 138; 43.5%) had more than 10 years of experience in the healthcare system, while 81 nurses (25.5%) had 5 years or less of professional experience. The most common age group was 26–35 years (n = 90).

The participants worked in a variety of settings: medical wards (n = 33, 10.4%), surgical wards (n = 7, 2.2%), intensive care units (n = 10, 3.2%), emergency departments (n = 14, 4.4%), and other settings (n = 262, 82.6%).

The age distribution was 19–25 years (n = 65, 20.5%), 26–35 years (n = 90, 28.4%), 36–45 years (n = 81, 25.6%), 46–55 years (n = 44, 13.9%), and ≥56 years (n = 37, 11.7%). See

Table 1. Baseline characteristics.

Measure		n	%
Sex	Male	61	19.24
	Female	256	80.76
Age (years)	19–25	65	20.50
	26–35	90	28.39
	36–45	81	25.55
	46–55	44	13.88
	>56	37	11.67
Education	High school	100	31.54
	Higher vocational	145	45.74
	University level	72	22.71
Years of working	Less than 1 year	37	11.67
	1–5 years	44	13.88
	6–10 years	56	17.66
	11–20 years	90	28.39
	21–30 years	48	15.14
	More than 30 years	42	13.25

for detailed baseline characteristics.

Knowledge about hypoglycemia was assessed using a structured questionnaire consisting of 26 scored items, covering symptoms, risk factors, medications, and emergency management protocols. Each correct answer was awarded one point, and the final score was scaled from 0 to 100.

The mean knowledge score was 66.9% (SD = 17.8), with a median of 65.4% (IQR = 53.9–84.6%). A total of 237 participants (74.8%) scored at or above 50%. The distribution of knowledge scores is shown in

Table 2. Descriptive statistics for knowledge and attitude scores among nurses.

Measure	Mean	Median	SD	Min	Max	n
Knowledge score (%)	66.9	65.4	17.8	23.1	88.5	317
Attitude score (1–5 scale)	3.42	3.67	0.70	2.00	4.83	317

; see

Table 3. Knowledge according to questions that test knowledge about hypoglycemia.

	Correct Answer		Wrong Answer
	n	%	n	%
What is the cut-off value of capillary blood glucose (mmol/L) for hypoglycemia?
<4 mmol/L	173	54.58	144	45.42
What are the risk factors for hypoglycemia?
Recovery from an acute illness	206	64.98	111	35.02
Insulin or oral hypoglycemia therapy at an inappropriate time in relation to a meal	271	85.49	46	14.51
List the symptoms of hypoglycemia.
Hunger	223	70.35	94	29.65
Trembling	284	89.59	33	10.41
Sweating	290	91.48	27	8.52
Annoyance	275	86.75	42	13.25
Headache	208	65.62	109	34.38
Convulsions	154	48.58	163	51.42
Which of the following diabetes medications can cause hypoglycemia?
Gliclazide	156	49.21	161	50.79
Insulin	276	87.07	41	12.93
15 g of fast-acting carbohydrates is equal to:
3 dextrose candy	163	51.42	154	48.58
Which of the following is a neuroglycopenia symptom of hypoglycemia?
Speech difficulties	220	69.40	97	30.60
Which of the following nutritional problems is a risk factor for developing hypoglycemia?
Changing the time of eating the main meal of the day	209	65.93	108	34.07
Long time of starvation	265	83.60	52	16.40
Inability to access the usual snack	191	60.25	126	39.75
Eating less carbohydrates than usual	179	56.47	138	43.53
What intervention should be performed during an episode of hypoglycemia in a patient who is conscious and oriented?
Give 200 mL of any juice	253	79.81	64	20.19
What interventions during an episode of hypoglycemia should be carried out in a patient who is unconscious or has convulsions?
Check the patency of the airway, breathing, and circulation	285	89.91	32	10.09
Administer 1 mg of glucagon intramuscularly/subcutaneously	255	80.44	62	19.56
When should blood glucose be rechecked after an episode of hypoglycemia?
15 min after therapy	200	63.09	117	36.91

for details about each question.

Attitudes toward the management of hypoglycemia were assessed using six Likert-scale items. Two items were reverse coded to ensure consistent polarity. Each item ranged from 1 (strongly disagree) to 5 (strongly agree), and a composite attitude score was calculated as the mean of the six items.

The internal consistency of the scale was acceptable (Cronbach’s alpha = 0.65). This moderate level of internal consistency suggests the need for psychometric refinement of the attitude scale. Future studies should consider conducting factor analysis or adapting items from validated attitude instruments in diabetes self-management contexts.

The overall attitude score had a mean of 3.42 (SD = 0.70), with a median of 3.67 (IQR = 2.83–3.83), indicating generally positive perceptions toward hypoglycemia management.

The participants showed generally positive attitudes toward hypoglycemia management (

Table 4. Composite and item-level results for attitudes toward hypoglycemia management.

Attitude Statement	Median	IQR
Patients with hypoglycemia receive insufficient care compared to hyperglycemia	3.00	2.00–4.00
Management should include a multidisciplinary approach	4.00	3.00–5.00
Prevention and proper treatment do not reduce hospital costs *	2.00	2.00–4.00
Strict glycemic control is associated with good outcomes	4.00	3.00–5.00
Standardized protocols are additional work for nurses *	2.00	2.00–4.00
Guidelines improve patient prognosis	4.00	3.00–5.00

* Note: Items were reverse scored for the composite score calculation.

). The strongest agreement (median 4.00, IQR 3.00–5.00) was seen for statements that management should include a multidisciplinary approach, that strict glycemic control is associated with good outcomes, and that guidelines improve patient prognosis. Conversely, the respondents tended to disagree (median 2.00, IQR 2.00–4.00) with the notions that prevention and proper treatment do not reduce hospital costs and that standardized protocols constitute additional work for nurses. The perception that patients with hypoglycemia receive insufficient care compared to those with hyperglycemia was neutral to slightly positive (median 3.00, IQR 2.00–4.00).

Differences in knowledge and attitudes were tested using non-parametric Mann–Whitney U tests, and standardized effect sizes (r) were calculated. Education level was grouped into secondary vs. tertiary (which includes both higher vocational and university degrees).

In

Table 5. Group differences in knowledge and attitude scores.

Comparison	U Value	p-Value	Effect Size (r)	Interpretation
Knowledge~Education	19,507.0	<0.001	0.641	Large effect
Attitude~Education	19,614.0	<0.001	0.649	Large effect
Knowledge~Experience	11,305.0	0.013	0.138	Small effect
Attitude~Experience	8441.5	0.700	0.023	No significant effect

the group differences in knowledge and attitudes are presented. Nurses with tertiary education scored significantly higher on knowledge (U = 19,507.0, p < 0.001, r = 0.641) and on attitude (U = 19,614.0, p < 0.001, r = 0.649) compared to nurses with secondary education. When comparing by experience (≤5 years vs. >5 years), nurses with more experience had significantly higher knowledge scores (U = 11,305.0, p = 0.0134, r = 0.138), but their attitudes did not differ significantly (U = 8441.5, p = 0.70, r = 0.023).

Figure 1 shows the relationship between the nurses’ percent-correct knowledge scores and years of experience (r = 0.138, p = 0.013).

3.2. Multivariable Analysis

To explore the independent effects of education, experience, sex, and workplace setting, we conducted multivariable logistic and linear regression analyses.

In the logistic model, the outcome was defined as having a ≥50% correct knowledge score. Tertiary education was the strongest independent predictor of sufficient knowledge (OR = 68.30, 95% CI: 19.92–234.21). Female sex was associated with higher odds of adequate knowledge (OR = 2.59, 95% CI: 1.02–6.62), while experience and workplace type were not statistically significant predictors.

In the linear regression model of attitudes, tertiary education remained a strong predictor of more favorable attitudes (β = +1.02, p < 0.001), while sex, experience, and workplace showed small or borderline effects, as shown in detail in

Table 6. Multivariable analysis of predictors of knowledge (logistic) and attitudes (linear).

Predictor	Knowledge ≥ 50% (OR [95% CI])	Attitude Score (β [95% CI])
Tertiary education	68.30 [19.92, 234.21]	+1.02 [0.90, 1.15]
>5 years experience	0.97 [0.38, 2.45]	–0.04 [–0.16, 0.10]
Female sex	2.59 [1.02, 6.62]	+0.13 [–0.02, 0.28]
Emergency workplace	4.78 [0.44, 52.11]	+0.26 [–0.03, 0.56]
Other workplaces	NS	NS

Note: OR = odds ratio; β = regression coefficient; NS = not significant.

.

4. Discussion

This study evaluated the knowledge and attitudes of nurses in Croatia regarding hypoglycemia management and found that while overall competence was satisfactory, significant disparities exist based on education level. Using robust statistical methods, including effect sizes and multivariable regression, we confirmed that tertiary education—defined as completion of a higher vocational or university-level nursing program—was the most consistent predictor of both adequate knowledge and more favorable attitudes. This reinforces the idea that formal educational pathways are not only gateways to licensure but may also serve as critical determinants of clinical preparedness.

Our findings are broadly consistent with international literature but also extend it in key aspects. A study conducted in Saudi Arabia [10] reported a marked discrepancy between perceived and actual diabetes-related knowledge among nurses, suggesting a tendency to overestimate one’s clinical competence. This gap was echoed in Jordan [11], where it was also found that nurses struggled particularly with applying theoretical knowledge in inpatient settings. Similarly, a study from the Philippines [13] reported an overall knowledge score of just 62.8%, with less than 18% of nurses correctly identifying the carbohydrate equivalence of hypoglycemia treatments.

Croatian nurses achieved a knowledge score of 66.9%, a value that sits squarely inside the 46.7–81.8% range reported across ten recent cross-sectional surveys of hypoglycemia or inpatient-diabetes knowledge worldwide. As in those studies, our weakest domains were neuro-glucogenic symptom recognition and carbohydrate dosing, reinforcing the idea that these are universal curricular blind-spots. Education level emerged as the dominant predictor of knowledge in Croatia—exactly mirroring patterns observed in Jordan, Poland, and the United States—whereas years of experience, ward type, and age showed inconsistent or null effects in both our data set and the wider literature [15,24,25,27,28,29,30,31].

The convergence of the findings strengthens external validity and indicates that upgrading pre-licensure and early-career training, rather than relying on experiential learning, is likely to yield the greatest competency gains. Notably, only a minority of published surveys have correlated attitudes with demographic factors; our demonstration that tertiary education also predicts more favorable attitudes therefore adds a novel dimension and underscores the importance of embedding evidence-based hypoglycemia protocols during formal education rather than leaving them to on-the-job diffusion. Finally, our CHERRIES-guided nationwide e-survey provides a rapid, scalable audit model that could be replicated in other countries to track progress following targeted educational interventions.

The role of sex as a predictor of knowledge—where female nurses outperformed males—warrants further exploration. It may reflect underlying differences in engagement with continuing education or in perceived roles within the care team.

Importantly, few studies in the field have used multivariable analysis to isolate independent predictors of knowledge and attitudes. Most rely on bivariate tests without adjustment for confounding variables. By employing regression models, this study provides stronger evidence for causally relevant factors, particularly the modifiability of education as a target for intervention.

These findings should also be interpreted in the context of nursing education structures. Croatia’s tiered educational model, with distinct tracks for secondary, higher vocational, and university-level nursing qualifications, bears similarities to systems in Germany and the Netherlands yet differs from North American models. Understanding how these educational pathways affect clinical competence across settings may offer insights into broader workforce planning, credentialing, and international harmonization of nursing standards.

These findings have practical implications. As Croatia continues to align with European nursing education standards, investment in tertiary nursing education and structured clinical training on diabetes management could significantly enhance patient safety.

Practical interventions should be guided by evidence from structured diabetes education programs. For instance, nurse-led models, such as the Hy-NEWSS protocol (Hypoglycemia—Nursing Education and Workflow Support System) or DSME (Diabetes Self-Management Education) have demonstrated success in enhancing nurse confidence and standardizing hypoglycemia care [32]. Adapting similar frameworks to the Croatian context could reinforce knowledge and translate into safer inpatient practice.

Moreover, standardized in-service training on hypoglycemia management, ideally integrated with clinical protocols and checklists, may offer a high-yield strategy to close the remaining competence gaps.

4.1. Limitations

The main drawback of this study is the small sample size. According to data from the Croatian Chamber of Nurses, on 24 April 2024, a total of 41,331 nurses were employed in the health system of the Republic of Croatia. This means that 0.7% of nurses in the Republic of Croatia participated in this research. However, it is necessary to consider the limitations of data collection through the Facebook group, because not all nurses in Croatia are members of that group. This method of data collection can affect the representativeness and generalizability of the research results. The group has 18,700 members, which means that the response rate among group members was 1.7%. It is possible that nurses were overwhelmed by invitations to complete online surveys, which may have contributed to the low response rate in this group. Future research should adopt multi-channel recruitment strategies beyond social media to improve sample diversity and representativeness. Potential avenues include institutional mailing lists, hospital intranet systems, national nursing associations, and academic nursing programs.

In addition, self-selection bias may have led to overrepresentation of nurses with greater interest or competence in diabetes care.

4.2. Ideas for Future Research

It is recommended to expand future research to include a larger number of respondents. Furthermore, it would be useful to classify the respondents according to the shifts they work in and compare their knowledge, given that episodes of hypoglycemia often occur at night.

5. Conclusions

This study shows that nurses in the Republic of Croatia possess generally strong knowledge and favorable attitudes regarding hypoglycemia management. Using a validated 26-item scoring system and a six-item attitude scale, we found that tertiary education emerged as the strongest independent predictor of both knowledge and attitudes. Regression models confirmed this relationship even after adjusting for experience, sex, and workplace setting. Future initiatives should prioritize competency-based training to ensure consistent, evidence-based hypoglycemia care across clinical settings.